Loading activity…
@moyix
Building offsec agents: http://xbow.com PGP http://keybase.io/moyix/
AI
CVE-2026-98432 challenged one user's AI-assisted exploit attempts; a reply on X claims Anthropic used an unreleased model.
AI
OpenAI's alleged plan prompted a critic on X to warn it could deter security researchers from using Daybreak.
AI
A post citing Anthropic says Claude's exploit ran commands as root against Linux servers with a Multipath TCP listener.
XAI
One post argues that stated limits contradict the first pillar; a reply reads the pillars as goals, not achievements.
AI
A user suggests steering one neuron up and another down, joking that the result would be “evil Qwen.”
AI
OpenAI says agent-led tests exposed VM escapes and limited internet routes, leading to patches and stronger monitoring.
AI
Another user says Claude's memory described their cat Katya as their wife.
AI
Two posts point to security incidents and urge AI-assisted defenses; one specifically calls for faster vulnerability detection and patching.
AI
The zero-loss weight changes are Litman's reported result; moyix questions his supply-chain zero-day framing.
AI
A user calls the result impressive but says the system is multiple times less efficient than Sol.
AI
A post says AI-agent inquiries are already bypassing barriers that once limited access to government and company services.
AI
A user welcomes the increase, saying strong incentives to hoard bugs can work against immediate reporting.
AI
XBOW Labs says Google awarded the chain the first $250,000 Chrome full-chain reward under rules introduced in April 2026.
AI
A user says the Claude app is finally “v dialed in,” alongside their praise for Opus 5.5.
AI
A user wonders whether reasoning in latent space, rather than executable chains of thought, might make a difference.
AI
Usage policy update adds new rules and highlights the potential for misuse.
AI
A user says they'd forgotten making the prediction and discussing why ExploitGym could be problematic.
AI
A user says they asked CVP to review a draft about their “full chain,” then wrote, “so much for CVP.”
AI
OpenAI says it drew on an independent mathematics advisory group’s advice about how to release the results.
AI
The user says they asked the AI to set up two sandboxed agents with a goal of finding a way to communicate.
